Unverified50% confidenceFactExact time
The largest NCM model the authors trained is under 1 billion parameters, and whether it holds at 27B or even larger scales has not been verified.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/28/2026
First Seen
Valid until: 1/26/2027
Sources
Related Entities
Related Claims
UnverifiedThe 29-million-parameter encoder trained by the researchers achieved a popularity-weighted cross-fold AUC of only 0.486 on 985 statements, indistinguishable from the corpus prior.68% similarUnverifiedThe 9-billion-parameter model in the paper lagged 6 points behind a clean summarization baseline before being trained with reinforcement learning.64% similarUnverifiedThe research team trained a series of decoder-only architecture models from scratch, with parameter scales ranging from 47 million to 349 million.63% similarUnverifiedTraining a language model from scratch under VRAM-constrained conditions typically involves choosing architectures with a smaller parameter count, usually on the order of 100 million to 500 million parameters.63% similarUnverified针对有限意图集合训练的轻量级专用模型参数量通常在10M以下，其推理延迟可比通用LLM调用低2-3个数量级63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1107922API
curl https://kongchang.com/api/v1/knowledge/claims/1107922MCP
get_claim(id=1107922)